Redacted for Privacy - Oregon State University

11 downloads 0 Views 2MB Size Report
Oct 21, 1993 - Date thesis is presented. October 21, 1993. Assembled by. Sue Pullen for. Douglas G. Pyle. Redacted for Privacy. Redacted for Privacy.
AN ABSTRACT OF THE THESIS OF

Douglas G. Pyle in

Oceanography

for the degree of Doctor of Philosophy presented on

October 21, 1993

Title: Geochemistry of Mid-Ocean Ridge Basalt Within and Surrounding the Australian-Antarctic Discordance

-

Redacted for Privacy Abstract Approved: David M. Christie

The Australian-Antarctic Discordance (AAD) is a unique feature of the global mid-

ocean ridge (MOR) spreading system located in the Southern Ocean between Australia and

Antarctica. Dramatic changes in MOR morphology and geochemistry, comparable to the contrasts between the slow-spreading Mid-Atlantic Ridge and the fast spreading East Pacific Rise, coincide with an abrupt axial depth transition between the AAD and adjacent

sections of the Southeast Indian Ridge (SEIR). A regionally uniform, intermediate spreading rate requires that mantle temperature and melt supply controls the physical and chemical characteristics of the spreading axis.

An isotopically defined boundary (Sr, Nd, and Pb) between Indian Ocean-type and Pacific Ocean-type MORB mantle lies beneath the AAD. The transition between these two

upper mantle reservoirs is abrupt, over 3000 km west of the AAD. The isotopic influence of these hot spots has been documented southeast of Amsterdam Island along the SEW [Dosso et aL, 1988], but MORB compositions apparently return to 'ambient' values within 150 km, implying a very

limited region of influence. The isotopic and topographic effects of large hot spots on spreading centers are well documented in other places, but the most prominent examples

extend only -1000 km in the case of Iceland and -500 km in the case of the Galapagos

[Hart et al., 1973; Sun et al., 1975; Schilling et al., 1982; Schilling, 1985]. The topographic gradient on either side of the AAD appears to be a consequence of the unusually deep axial bathymetiy of the AAD rather than directly related to any hot spot, and

as we have shown, the expected isotopic gradients are absent.

Pacific mantle migration: Alvarez [1982, 1990] proposed that Pacific upper mantle

must necessarily migrate through gaps between deep continental roots surrounding the Pacific basin as the long-term shrinkage of the Pacific Ocean proceeds. In terms of this hypothesis, the isotopic boundaiy in the AAD would represent the leading edge of Pacific

mantle which began to flow westward when a gap between the continental roots of Australia and Antarctica opened through the South Tasman Rise. Presumably, 'Indian'

upper mantle existed beneath the region prior to the inflow of Pacific mantle, having migrated eastward during the early stages of continental rifting which propagated from west

to east [Mutter et al., 1985, Royer and Sandwell, 1989; Veevers and Li, 1991]. The westward mantle migration rate of 25 mm/yr estimated from the isotopic data within the AAD would place the opening of the South Tasman Rise to mantle

flow

at -50 Ma, broadly

consistent with the opening of the circum-Antarctic region to oceanic circulation at 35 to 42

Ma [Royer and Sandwell, 1989; Veevers et al., 1991; Hinz et aL, 1991]. This hypothesis implies that the isotopic boundary has only recently arrived beneath the AAD and it is not directly related to the long-lived geophysical and morphologic features of the region.

Passive Mantle Flow Implicit in each of the active flow hypotheses is the concept that the upper mantle beneath the AAD is cold because it represents the distal regions of a horizontal mantle flow

regime. It is also possible that the characteristics of cold mantle beneath the AAD may generate or reinforce a passive lateral mantle

flow

into the region.

Mantle coldspot: Hayes [1976] and Anderson et al. [1980] suggested that the AAD

is associated with a mantle "coldspot". This concept is not clearly defined but it involves,

41

in some sense, the downwelling inverse of a hot spot. Fluid dynamic modeling of convection in an internally heated mantle [Bercovici, 1989] appears to favor focused upwelling and more diffuse, sheet-like downwelling, suggesting that a down-flowing coldspot is unlikely to be a stable feature. In any case, as Marks et al. [1990, 1991] have shown, the V-shaped trace of the AAD and the northward migration of the SEIR require that the physical processes producing the AAD cannot be associated with a point source or sink fixed in the mantle reference frame.

Passive mantle flow: Forsyth et al. [1987] proposed a self-consistent model on the

basis of seismic data that showed the AAD to be underlain by cooler-than-normal mantle

and that the lithosphere thickens away from the AAD spreading axis more rapidly than

average. Cool mantle, being closer to its solidus, would produce less melt and overall thinner crust. Cooler and/or volatile-depleted asthenosphere beneath the AAD should also

be more viscous than usual [Forsyth et al., 1987]. As a result, lateral return flow due to

plate separation along the SEIR may draw material along-axis from 'normalt (i.e. less

viscous) asthenosphere east and west of the AAD. This model does not require either mantle downwelling beneath the AAD or a specific external driving force for subaxial flow,

apart from plate separation. This model is consistent with many of the known features of

the AAD and with the presence of propagating rifts [Phipps Morgan and Parmentier, 1985]. It predicts that the isotopic boundary is a long-term (if oscillatory) feature of the AAD, but it does not provide an explanation for the existence of cold mantle in the first place.

Implications for mantle dynamics beneath the AAD Mantle flow toward the AAD has been inferred from the convergence of propagating rifts from the east and west during the last 25 Ma [Marks et al., 1990, 1991; Vogt et al., 1983] and from the pronounced bathymetric gradients along the SEIR [Weissel

and Hayes, 1974; Hayes and Conolly, 1972]. Fundamental questions as to whether mantle

flow towards the AAD is active or passive and whether mantle downwelling is occurring

42

remain unresolved. Tangible evidence for recent mantle flow beneath the AAD is provided by the off-axis data from segment B5.

Whether the migration of the isotopic boundary represents a continuous, long-term

westward flow of Pacific mantle or a recent perturbation of the boundary about a mean position beneath the easternmost AAD is not known. If the westward movement of the isotopic boundary across B5 has been a continuous phenomenon, then mantle migration

rates between 10 mm/yr and 40 mm/yr are permitted by existing data, depending on the exact ages of off-axis samples and assumptions about present (and past) positions of the isotopic boundary. We have assumed that the isotopic boundary coincided with the eastern

AAD bounding transform at 4 Ma and now lies at the B4/B5 transform, requiring Pacific

mantle flow of -25 mm/yr westward beneath the B5 spreading segment. The most interesting implication of steady, westward, Pacific mantle flow is that the isotopic boundary has only recently arrived beneath the AAD and that it is not coupled to, or formed

by, the mantle dynamics creating the regional gravity and bathymetric anomalies. Although

the timing of the arrival seems fortuitous, the consistency between inferred rates of migration and the likely timing for rifting of the South Tasman Rise allowing upper mantle

outflow from a shrinking Pacific basin [Alvarez, 1982, 1990] requires that this hypothesis be carefully considered.

An isotopic boundary that is instead related to the mantle dynamics creating the

AAD is equally plausible given the present sampling distribution. Movement of the isotopic boundary may simply reflect small-scale oscillations in mantle flow from the east and west. If small scale fluctuations in the boundary position have occurred throughout the existence of the AAI) depth anomaly (over the last 100 Ma), a much wider zone of mixing

and isotopic exchange might be expected to have developed between the two MORB

sources [Hofmann and McKenzie, 1985], leading to smooth gradients in all isotopic systems on either side of a median boundary position. The only evidence for this type of

source mingling is near the leading edge of the Pacific' source and, significantly, the B4

43

segment lavas lie well within the Indian MORB field on all diagrams. Furthermore, sharp

discontinuities in Pb and Nd isotope ratios remain between the B4E and B5W segments. At present, we believe that these observations imply that Pacific' mantle has not penetrated beyond the B41B5 transform and that there is a gradually diminishing 'memory' of 'Indian'

mantle incorporated into 'Pacific' lavas beneath B5, behind the migrating isotopic boundary. The present position and the abrupt nature of the eastern Indian MORB boundary at

-126°E is not easily explained by simple dispersion of low 206Pbf2Pb and Indian Ocean hot spot contaminants by channelized, sub-axial, hot spot-induced mantle flow, because the

AAD is beyond a reasonable distance over which hot spots are likely to have a significant influence. Dispersion of the Indian Ocean upper mantle is more consistent with a regional,

eastward flow of mantle towards the AAD in response to the separation of Australia and

Antarctica. By this mechanism, isolated 'Kerguelen'- and 'Reunion'-type contaminants

could be drawn eastward with depleted Indian MORB mantle until it is halted by the

countervailing westward flow of 'Pacific' mantle. The unusually abrupt nature of the eastern Indian MORB boundary may reflect an active, westward Pacific mantle flow, driven by the shrinkage of the Pacific basin [Alvarez, 1982, 1990], which is colliding with a more passive Indian Ocean mantle. The western boundary to the Indian MORB reservoir is not as sharp or well-defined [Mahoney et al., 1992] because both the Atlantic and Indian

basins are presently growing and no flow of upper mantle from one basin to another is required to accommodate the continental motions.

SUMMARY The new results presented here, together with previous data [Klein et al., 1988], show that the isotopic boundary between Pacific and Indian MORB mantle provinces is

remarkably abrupt beneath the AAD. No other boundary of comparable abruptness and contrast has been documented anywhere along the global mid-ocean ridge system. The transition between Indian and Pacific MORB sources occurs within a narrow zone (-40 km) beneath the spreading axis of a single segment (B5) within the eastern AAD. Limited

off-axis sampling from this segment indicates that the boundary has migrated approximately 100 km westward since 4 Ma. Despite the movement of Pacific mantle

beneath the AAD, the 'Indian' and 'Pacific' sources have maintained distinct isotopic signatures, apart from the isotopic gradients along the B5W segment. The characteristics of the Pacific-Indian upper mantle transition suggest that the isotopic boundary is near the

B4/B5 transform and that a 'memory' of the displaced 'Indian' mantle signature is incorporated into 'Pacific' lavas behind the leading edge of a migrating Pacific mantle.

Two classes of upper mantle flow towards the AAD can explain the geophysical and isotopic characteristics of the AAD: (1) sub-axial mantle flow in response to spreading over a cool mantle, and perhaps to along-axis topographic gradients; (2) long-range inflow

of upper mantle from a shrinking Pacific Ocean in which Pacific mantle is gradually displacing Indian Ocean mantle. In neither case are the origin of the cold mantle beneath the AAD and the depth and gravity anomalies centered on the AAD satisfactorily explained.

The first model predicts that the isotopic boundary is linked to the long-lived geophysical anomalies and morphological features which define the AAD, whereas the second model requires that they are independent and that the isotopic boundary has only recently arrived

beneath the AAD. The present sample distribution does not allow an unequivocal discrimination between these two models; however, both lead to specific, testable predictions as to the off-axis configuration of the isotopic boundary. The isotopic data are overwhelmingly biased to 'zero-age', spreading axis samples and it is clear from this study

45

that significant information about mantle flow in the region, and therefore the mantle dynamics beneath the AAD, is to be found off-axis.

Chapter 3

Geochemical and Morphological Contrasts in the Vicinity of the Australian-Antarctic Discordance

D. G. Pyle and D. M. Christie

ABSTRACT The Southeast Indian Ridge (SEIR) in the vicinity of the Australian-Antarctic Discordance (AAD) is a region of unusually diverse spreading axis morphology and midocean ridge basalt (MORB) geochemistry. Distinct changes in major and trace element

composition are shown to coincide with the abrupt change in axial segmentation and morphology that accompanies the transition between the AAD and SEIR spreading segments to the east (i.e., Zone A). Since spreading rate is uniform throughout this region, compositional and morphological variations are most likely related to variations in melt

production beneath the spreading axis. The AAD spreading system is melt deficient and compositional variations within the AAD MORB suite are primarily controlled by partial

melting of the mantle source. The Zone A spreading system is melt saturated and compositional variations are dominated by shallow pressure, crystal fractionation. The major element characteristics of the AAD indicate that these basalts are derived

from low average pressures of melting and low degrees of melting. The Zone A basalts are derived from higher average pressures and higher average degrees of melting. The major element contrasts between these two MORB suites are consistent with a variations in mantle temperature producing variations in mantle melting. However, trace element systematics show that lavas produced by low degree melting throughout the region (both AAD and Zone A) have very similar source mantle signatures, but retain very different major element

compositions. Differences in partial melting alone can not produce these relationships unless trace element heterogeneity in the source mantle is evoked. An alternative model is presented in which the modal proportion of clinopyroxene (cpx) in the source mantle is varied. This model satisfies the trace element conditions and predicts that the AAD mantle

is higher in modal cpx. Therefore, AAD compositions result from lower degrees of melting from a more fertile source mantle (i.e., higher proportion of cpx).

47

INTRODUCTION The formation of oceanic crust at mid-ocean spreading centers is controlled by the combined effects of spreading rate and mantle upwelling. The interaction between these two processes results in a spectrum of petrogenetic and tectonic conditions along-axis that

are recorded by axial morphology and mid-ocean ridge basalt (MORB) composition. In

this study, we present major and trace element data for basaltic glass samples dredged along 400 km of the eastern Southeast Indian Ridge (SEIR), an area encompassing the eastern three-quarters of the Australian-Antarctic Discordance and adjacent spreading centers to the east. Along this portion of the SEIR, an unusually diverse suite of MORB compositions has erupted from spreading centers which display, at a uniform, intermediate

spreading rate (74 mm/yr full rate) [De Mets et al., 1990], nearly the entire range in axial morphology observed at slow and fast spreading systems. This requires that spreading rate is not the dominant variable for controlling either axial morphology or basalt petrogenesis.

More likely, variations in mantle upwelling patterns related to variations in mantle temperature dominate spreading conditions in this region [West et al., 1994; Christie et al.,

1988; Christie et al., 1990]. The SEIR within and adjacent to the AAD provides an uncommon opportunity to examine the effect of mantle upwelling on melt production, MORB petrogenesis, and spreading axis morphology independent of spreading rate.

BACKGROUND The Australian-Antarctic Discordance (AAD) (Hayes and Conolly, 1972) is a 500 km

section of the eastern SEIR characterized by a pronounced, along-axis change in segmentation patterns and unusually deep axial bathymetry relative to adjacent SEIR spreading centers to the east and west (commonly referred to as Zone A and Zone C, respectively; Fig. ifi. 1). Within the AAD, spreading centers are relatively short ( 8.0% MgO presented in the paper (see Fig. 111.4). The resulting fractionation corrections are Fe8 0=Fe203+ 1.923 * (MgO-8.0) for the Zone

A suite and Fe8 0=Fe203+0.691 * (MgO-8.0) for the AAD suite. The Fe80 formulations outlined here do not modify the interpretations obtained using conventional corrections slopes of m= 1.664 for Zone A and m=.924 for

AAD (note .924 is the mid-Cayman Rise regression and should be appropriate for the AAD suite) [Klein and Langmuir, 1989}. However, the conventional fractionation corrections increase the scatter on the diagram

and amplify "local-trend" relationships. Tig0 values are calculated using conventional algorithms. No corrections have been applied to samples with greater than 8.0% MgO.

84

AAD

zone C

zone A

(zone B)

+

Na8.O B3

low% melt

I-

B5axis

D

0 B5 off-axis

(>A3 A A2

B4,B3,B2

iAl

cayman field

3.25

MAR 1l°N Kane

B5

+

+

0 /

+/ £__ 2.75

/

:

I

B5

global 2.5

A

trend

:

Famous Amar

.:

Al High% melt

A2 2.25

I

1

7

7.5

8

I

I

I

8.5

9

I

I

I

I

I

I

I

9.5

10

10.5

Fe8.O 3.75

3.5

PRT lavas

3.25

+%

3

A

A61

2.75

2.5

n

A

Ti80 2.25 0.8

1

Figure 111.7

1.2

1.4

1.6

1.8

2

2.2

environments appears to be similar. Melts are generated and quickly erupt to the surface

without encountering melts from other parts of the melting column. The variety in REE patterns for lavas from these environments supports this conclusion.

Niu and Batiza [1991] extended the fractionation correction concept to other major elements and derived formulations to quantify MORB composition to physical parameters controlling melt generation within the mantle. Applying these formulations to the AAD and

Zone A data, the relative differences in the melt regimes can be assessed (Table 111.7).

Melts within the AAD melting column traverses a more restricted pressure interval, although the final pressure of equilibration is higher and the mean pressure of melting is

lower than for Zone A melts. Zone A melts are generated over a much greater depth interval, with the final pressure of equilibration being lower and the mean pressure of melting being higher than that of the AAD. The values of P0 appear to be a fairly good estimate of the base of the melt column, however, the high final melt pressures (Pf) seem

unusually high, possibly an artifact of the decompression melting model. The high Pf estimates might imply that decompression melting of a single mantle source composition

does not occur within an individual melting column, particularly if the melting column spans a considerable pressure range.

Table ffl.7 Na80 Zone A AAD

2.75 3.50

Fe80 10.5 8.5

Si80

Ca80

49.5 51.0

11.5 10.5

Al80 15.2 16.2

P0

Pf

F

(kb)

(kb) 14.0 8.2

(%)

20.4 9.7

16.7 11.2

T0 (C°) 1413 1283

Tf (C°) 1291 1225

Estimated X8.O values have been visually determined (not calculated) from MgO variation diagrams (Fig.ffl.4) and represent an 'average composition for each region at 8.0% MgO. The actual values depend on several assumptions including the experiments used to parameterize the model (Niu and Batiza, 1991) and the starting mantle composition (in this case MPY-40; Falloon et al., 1987). The relative differences are much more important than the actual values.

Parental Melts and Differentiation Direct comparison of primitive MORB compositions is the rationale behind correcting

MORB to an 'unfractionated' MgO value of 8.0 %. The identification Zone A and AAD primary or parental melts, and their differentiation trends is useful for understanding the melting processes leading to their formation. Glasses with greater than 8.0 % MgO were

recovered from both the AAD and Zone A allowing a straight forward comparison of primitive melt characteristics within each region. A modified version of the one atmosphere

fractionation model of Nielsen [1985] (Fig. 111.4) and the pseudoternary projection of Grove et al.[1992] (Fig. 111.8) have been used to evaluate potential parental melts and fractionation histories.

The primitive nature of the AAD MORB suite as a whole, suggests that only olivine

removal has affected their compositions. Several AAD glasses could be considered parental but no evidence exists to suggest one parent is common to the entire AAD suite.

The diversity of the AAD lavas cannot have been produced by any conceivable low pressure fractionation trend, as is clear from large variations in CaO/Al203 and Na20 at constant MgO values and a wide array of AAD REE patterns. Projection from plagioclase

suggests that AAD lavas last equilibrated between 2-6 kb in the mantle, never reaching a low pressure oI-pl-cpx cotectic (Fig. 111.8). Position within the plagioclase projection is unrelated to mg#, suggesting that high pressure fractionation is not a significant process in the differentiation of AAD lavas, although a limited influence can not be entirely ruled out for samples from dredge MW18.

Variations in CaO/A1203 between .70 and .80 over a wide range in MgO, and the sub-parallel REE patterns of Zone A lavas, indicate low-pressure fractionation is a major factor in the differentiation of this MORB suite. Several high-MgO, zone A glasses were evaluated as possible parental melts, although none of these 'parental' compositions satisfy

all of the major element variations by one-atmosphere fractional crystallization modelling

(Fig. 111.4). MW17-26 is the most likely high-MgO parent since its REE patterns most

closely resemble those of more evolved zone A lavas. The modelled one atmosphere fractionation path of MW17-26 predicts slightly lower Fe203 and Ti02, and considerably

higher Na20 and Si02 than is observed in evolved Zone A lavas. High Na20 and low CaOIAl2O3 is a persistent problem if primitive Zone A are indeed parental to more evolved

Zone A compositions. Similar discrepancies between predicted fractionation trends of primitive zone A lavas and their evolved counterparts are observed irrespective of the low pressure fractionation model (e.g., Nielsen [1985], Weaver and Langmuir [1990]) This is

somewhat unexpected given the coherent, 'low' pressure fractionation appearance of the major element data and the REE pattern similarities between MW 17-26 and evolved Zone A

basalts. If only Zone A data with 80 km [Salters and Hart,

1989; Beattie, 1993]. Both Sc and heavy rare earth elements have high partition coefficients for garnet and melting within the garnet stability field will buffer the abundance

of these elements until garnet is no longer a residual phase. Until garnet is completely

105

consumed, increasing Sc and heavy rare earth element abundances will reflect increasing

amounts of melting. Accepting that Zone A major element variations require a deeper

melting column, a greater influence of garnet on Zone A melts might be predicted. If melting begins in the garnet stability field for Zone A lavas, but not for AAD lavas, then Sc and heavy rare earth element abundances would be expected to be lower, rather than higher

in Zone A lavas. If residual garnet affects both the Zone A and AAD lavas, then garnet must be more depleted in the Zone A melt column, or completely consumed, relative to the

AAD to produce a high heavy rare earth element signature in zone A lavas. Furthermore, a greater degree of light to heavy rare earth element fractionation might be expected if garnet played a large role. Initiating melting within the garnet stability field for both the AAD and

Zone A would suggest very similar melt column configurations for both regions which is not apparent in the major element systematics of the two MORB suites. If melting begins at

pressures of garnet stability for MORB in general, then decompression melting and the progressively depletion of a self-contained mantle diapir from these depths seems highly improbable. Whether a garnet bearing mantle is incorporated into a melting model or not,

the modal mineralogy of mantle entering the melt regime beneath the ridge must vary between the AAD and Zone A.

106

CONCLUSIONS The unusual diversity of MORB compositions and spreading axis morphologies, at a

uniform spreading rate, are highly significant features of the SEIR in the vicinity of the AAD in terms of global accretionary processes. Magma processing after melt generation and prior to seafloor eruption appears to be fundamentally different between the AAD and

Zone A, resulting two distinct MORB suites with different styles of differentiation. The broad axial rise of the Zone A segments suggests that relatively large volumes of melt are

delivered to the neovolcanic zone. This 'excess' magma supply enables a low pressure magma system to develop, moderating the diversity of certain major and trace elements and

allowing for extensive fractional crystallization. In general, Zone A melts are produced over a greater depth range within the melting column and pooled beneath the ridge. Within

the AAD, the deep axial valleys suggest a 'deficient' magma supply. The primitive nature

of AAD lavas suggests that little, if any, low pressure fractionation has modified their composition and, consequently, much of the chemical diversity within the AAD is due to melting within the mantle source. AAD lavas are produced over a more limited, shallower depth range and rapidly erupted at surface with litfie modification.

Contrasts in axial morphology and MORB composition between the AAD and Zone A vary together. Therefore, the geochemistry and morphology of the ridge is controlled by

magma supply and magma supply is intimately tied to the underlying mantle temperature

[Klein and Langmuir, 1987; 1991; Forsyth, 1992; Sempéré et al., 1991]. Higher mantle temperatures produce enough melt to create an axial rise morphology and sustain a well

developed, low pressure magma system. The cooler mantle beneath the AAD does not allow low pressure fractionation and magma chemistry is controlled primarily by melting in

the source. The average degree of melting and subsequent differentiation history of the AAD and Zone A MORB suites is tied to thermal energy available to create and maintain a shallow level magma system.

107

As concluded by previous studies, the major element characteristics of AAD lavas are consistent with lower degrees of partial melting at lower mean pressures of melting relative

to Zone A lavas [Klein et al., 1991]. However, higher concentrations of moderately incompatible elements in Zone A lavas are inconsistent with the higher degrees of melting

suggested by parameters such as Na80. Therefore, differences in mantle melting beneath

Zone A and the AAD cannot be the sole cause of the trace element contrasts between primitive melts from both regions.

A mantle source common to both the AAD and Zone A is suggested by trace element

similarities between low degree melts throughout the region. In particular, the primitive Zone A propagating rift tip lavas share many trace element signatures of AAD lavas. The

AAD and Zone A propagating rift lavas differ in their major element characteristics, suggesting that this common trace element signature is derived from deeper depths beneath

the Zone A mantle. If melting is conceptualized as a column, then the fertile, undepleted base of this mantle melting column lies at a deeper depth below Zone A than it does below the AAD.

The condradictory melting information coveyed by the major and trace element

variations between the AAD and Zone A can be reconciled by varying the modal mineralogy of the mantle source beneath these regions. The AAD mantle has a high proportion of cpx (i.e., more fertile), relative to the total melting column, than does Zone A

mantle. This model satisfies major element requirements of decreasing melting toward the AAD, trace element signatures that indicate a common mantle source throughout the region,

and trace element abundances that are higher than expected in Zone A. Lower modal cpx in the mantle beneath Zone A suggests incorporation of more depleted mantle at depths above

the base of the melting column perhaps related to lateral mantle flow. This supposition follows from the similarity in trace element systematics of propagating rift tip and AAD lavas which indicate that fertile mantle exists beneath both regions, although it resides at higher pressures beneath Zone A. Alternatively, the mantle trace element composition must

be varied quite selectively in order to reconcile the trace element systematics with the major element systematics of the two MORB suites.

109

Chapter 4 Geochemistry and Geochronology of Ancient Southeast Indian and Southwest Pacific Seafloor

D. G. Pyle, D.M. Christie, J.J. Mahoney and R.A. Duncan

ABSTRACT An isotopically defined boundary between the Pacific Ocean and Indian Ocean MORB presently exists at the center of the Southern Ocean between Australia and Antarctica within the Australian-Antarctic Discordance (AAD). Isotopic contrasts between

axis and off-axis samples from the easternmost AAD spreading segment suggests that the boundary for these MORB sources has migrated westward at rate of -25 mm/yr during the last 4 m.y. The trace element and Sr, Nd, Pb isotope analyses of DSDP samples from sites

264, 265, 266, 274, 278, 279A, 280A, 282, and 283, coupled with 40Ar-39Ar age determinations of selected samples, are presented to determine the regional pattern of mantle source and evaluate the large-scale migration of the isotopic boundary over the last

60 m.y. The peripheral disthbution of these DSDP sites to the Southern Ocean basin records important regional variations in upper mantle source composition, however, the large-scale westward flow of Pacific MORB mantle since rifting of the South Tasman Rise at --4OMa cannot be resolved. DSDP basalts east of the South Tasman Rise have 87Sr/86Sr (0.70250.7029) and 206Pb/2Pb (18.80-19.48) characteristics that indicate a Pacific MORB-type mantle source beneath this region since seafloor spreading began in the Tasman Basin (-80 Ma). DSDP samples west of the AAD have Indian MORB-type, low 2Pb/'Pb and high 87Sr/86Sr characteristics. The South Tasman Rise basalts from DSDP sites 282 and 280A erupted between 60-69 Ma from an upper mantle source lower in 206Pb/2Pb than mantle farther east. Elevated 87Sr/865r, 208Pb/204Pb and 207Pb/204Pb values indicate an Indiantype upper mantle source for site 280A, but site 282 basalt is borderline between Pacifictype and true Indian-type mantle in this region.

110

INTRODUCTION

The Australian-Antarctic Discordance (AAD) In the center of the Southern Ocean between Australia and Antarctica lies the Australian-Antarctic Discordance (Fig. IV. 1). The AAD has long been recognized as an unusual section of the global spreading system because of its deep axial bathymetry (4-5

km), rough (and in places chaotic) topography, low gravity signal, high upper mantle

seismic wave velocities, and intermittent asymmetric spreading history [Weissel and Hayes, 1971, 1974; Forsyth et al., 1987; Marks et al., 1990; Sempéré et al., 1991; Palmer et al., 1991]. A history of multiple ridge propagation episodes towards the AAD from both

the east and west suggests convergence of upper mantle towards this region [Vogt et al., 1984; Phipps-Morgan et al., 1988].

Coincident with the distinct geophysical features of the AAD is an equally distinct geochemical boundary between isotopically defined Indian Ocean-type and Pacific Ocean-

type MORB [Klein et al., 1988]. All spreading axis samples dredged west of -126° E have

Indian Ocean MORB isotopic affinities and those dredged to the east have Pacific MORB

isotopic affinities (Fig. IV.2). The isotopic boundary is defined by an abrupt decrease in 206Pb/204Pb and 208Pb/204Pb, an increase in 87Sr/86Sr as the MORB source changes from

Pacific Ocean-type to Indian Ocean-type mantle (Fig. P1.3). Its is accompanied by a more

gradual declines in 207Pb/204Pb and 143NdJ1Nd. The isotopic transition is unusually

narrow, occurring over ':1p \--flM Is.-

.4/-./-.. KP'-

i

61J1

E/ f 267 I

\j-'

,.

___-.? -

____-

'II,,

70°

Figure IV.1

=.;ç.-\

. -5-------

. 600

- I

'1-'-.-

aiL I. IA Afl

I

,,.'

j.-

278

\

..'\

%S--\ \ - -

II-., II

.

_..__.________-.

-.5

I

t2'

114

Figure IV.2 Schematic summary of the SEIR tectonics within the eastern AAD and western zone A. Filled circles are MW8801 and Vema dredge sites that have Indian MORB isotopic characteristics and open circles are those that have

Pacific MORB isotopic characteristics. The isotopic boundary is presently near the transform boundary at 126° E. The grey line shows a hypothetical trace of the boundary assuming Pacific mantle migration began when the South Tasman Rise rifted.

115

130°E

128°E

126°E I

Sb

j1i

5:

1 I .....................................................................................

I

AAD

46°S

I.

zoneA

I

I

I

S

I

I

I

I___....___..__I

3

I

I

I__

I

'Indian'

mantle

Hypothetical Isotopic Boundary

Mior2tirnTr

Imn Ocean upper mantle Figure IV.2

Pacfic Ocean upper

0

48 S

116

Figure IV.3 Along-axis profiles of isotopic ratios from the SEIR (1 15°-138°). Open symbols represent Pacific' group compositions and filled symbols represent

'Indian' group compositions [Klein et al., 1988; Pyle et al., 1992]. Filled squares with open circles in B5 are off-axis dredges. Horizontal axis is distance (km) from the eastern bounding transform of the AAD. The present location of the isotope boundaiy is at or close to the B41B5 transform at 126°

E. Vertical solid lines show the position of transforms and vertical dashed lines show axial non-transform discontinuities and propagating rift tips.

zone

AAD (zone B)

C

zone

zone

AAD

zone

A

C

(zone B)

A

U. IUOO

15.60

0.7036 0.7034 0.7032 15.50

0.7030 0.7028

0.7026 i c An

0,5 132

38.50

0.5131

38.00 0.5 130

0.5129

37.50

00

19.50

3000 19.00

(m) 3500

. .11

I.

.

18.50

4000 4500

18.00

5000 17.50

S

I

\J.j27.s° E

I

axial

40 km

depth

transition zone

/

1-h

'Indian' MORB

550(1

-1000

-800

-600

-400

-200

0

200

400

600

800 km

°E

zoneA

Figure JV.3

zoneD

-800

-600

-400

-200

'Pacific' MORB

0

200

400

600

800 km

120 122 124 126 128 0f

zoneR

zoneA

.

126

-1000

114 116 118

114 116 118 120 122 124 126 128

PR

PR

I

128 130 132 134 136 138

126

128 130 132 134 136 138

118

DSDP Legs 28 and 29 is presented below to determine the regional implications for mantle flow in this region.

ANALYTICAL METHODS Glass and whole rock samples from ten DSDP sites (Table IV. 1) were analyzed for

major and trace elements by combined microprobe, X-ray fluorescence spectrometry (XRF), and inductively coupled plasma mass spectrometry (ICP-MS). Petrographic descriptions of the core samples condensed from the DSDP Initial Reports volumes for Legs 28 and 29 are summarized in Appendix 1. All basalt core samples were trimmed, and

sawn surfaces were ground on a lap wheel to remove contaminants, obtaining the freshest

possible material. These rocks were crushed in ceramic jaw crusher, washed in Millipore filtered water, dried, and hand-picked to remove alteration veins and vug material. Whole

rock powders were made using a tungsten carbide ring mill at Washington State University. All analyses are on splits of the same sample powder. Glass samples were crushed in a ceramic mortar, sieved to 30-mesh size fraction, and hand-picked to remove alteration and phenoctyst phases.

The major element concentrations of all glass samples were determined using a Cameca SX-50, four-spectrometer electron-microprobe at Oregon State University. Each

reported analysis represents the average of 5 spot analyses that have been normalized to standard BASL glass run along with the unknowns. Whole rock samples were analyzed

for major and trace elements (Ni, Cr, Sc, V, Ba, Rb, Sr, Zr, Y, Nb, Ga, Cu, Zr) by XRF at Washington State University following standard methods (Tables P1.2, IV.3, P1.4).

Further trace element analyses were completed by ICP-MS using a Fisons PQ2+ PlasmaQuad at Oregon State University for Sc, V, Cr, Ni, Cu, Zn, Rb, Sr, Y, Zr, Nb, Cs,

Ba, La,Ce, Pr, Nd, Sm, Eu, Gd, Th, Dy, Ho, Er, Tm, Yb, Lu, Hf, Ta, Th, and U (Tables

P1.2, IV.3, IV.4). The XRF and ICP-MS trace element results show good agreement except for low concentration elements (Table IV.5). At low concentrations, the ICP-MS values are assumed to be better (e.g., Rb, Ba).

119

Table IV. 1 DSDP sites sampled for this study. Site

LEG 28

LEG 29

264 265 266 267 274 278 279A 280A 282 283

Latitude (S)

Longitude (E)

34° 58.13' 53° 32.45' 56° 24.13' 59° 15.74

112° 2.68' 109° 32.45' 110° 6.70' 104° 29.30'

68°59.81'

173°25.64' 160° 4.29' 162° 38.10'

56° 33.42' 51° 20.14' 48° 57.44' 42° 14.76' 43° 54.60'

Magnetic --Aget Anomaly (Ma)

5b 6 15 13 12

147° 14.08

143°29.18' 154° 16.96'

t Kent and Gradstein (1986) magnetic anomaly time-scale

31

15

23 38 36 33

Core Rec.

Rec.

(m)

%

17 14 14

6 11

23 ?47 ?55

5 5 15

69

4

20 14

16 34 50 70 100 50 43

120

Table IV.2 Major and trace element composition of DSDP Leg 28 basaltic basement cores.

Na20 K20 P205

DSDP 274 DSDP 267 DSDP 265 DSDP 266 264-15 265-17-01 265-18-01 266-23-01 266-23-01 267-07-01 267-07-01 274-44-02 274-45-02 110-114 61-64 88-93 78-82 76-82 61-63 41-49 pc.5 cc w/r w/r w/r glass wlr w/r glass wir glass xrt xrt xrf probe probe xii probe xi xxl 52.00 51.11 51.88 50.95 50.32 51.05 50.07 49.51 55.08 1.71 2.01 1.31 1.32 2.18 2.28 1.58 1.63 1.53 17.82 17.05 14.62 15.80 14.77 17.14 16.67 13.92 17.32 9.64 8.73 9.17 8.98 11.62 7.50 10.92 7.87 9.62 0.15 0.11 0.17 0.15 0.14 0.20 0.19 0.12 0.11 5.17 6.05 7.86 7.87 7.18 6.37 8.36 9.40 5.22 7.87 11.00 1239 12.31 10.40 10.35 10.51 11.12 6.88 3.40 2.95 3.35 2.63 3.11 3.27 3.45 2.96 3.00 0.12 1.27 0.13 0.31 0.35 0.64 0.25 0.60 1.17 0.47 0.12 0.15 0.13 0.26 0.27 0.30 0.25 0.19

total

100.11

98.86

101.29

98.52

100.42

99.16

101.45

99.75

99.88

Sc V

27.9 215

29.2

116 19 16

42.7 372 285

38.8 336 201 90 56 110.8 4.5 129

42.9 292 296

41.4 260 259

46.8

Cr

33.3 232 325

38.4 267 207

79 70 75.2

78 77 75.5

DSDP 264

core interval split analysis S102

Tl02 Al203 FeOt MnO MgO CaO

ICP-MS (ppm)

Ni

Cu Zn Rb

Sr Y

Zr Nb Cs Ba

La Ce

Pr Nd Sm

Eu Gd

Tb Dy Ho

Er Tm Yb Lu Hf Ta Th U

90.9 20.2 205 25 120 7.2 0.10

187

50 81.4 9.2 297 29 139 14.4 0.11

104 9.91

307 19.47 43.52 5.13 20.78 4.81

24.14 3.24 14.69 3.96

1.53

1.41

5.09 0.81 4.79

4.37 0.75 4.63 0.93 2.75

0.91 2.51

0.36 2.10 0.30 3.25 0.53 3.72 0.42

Pb*

* isotope dilution analysis

0.41

2.55 0.38 2.86 0.91 1.03 0.31 0.700

186

303 268 54 79.9 7.2 257 25 112 13.2

0.44 97 8.60 20.49 2.77 12.80 3.41 1.23 3.92 0.67

4.04 0.83 2.49 0.37 2.22 0.34 2.41

2.79 1.34 0.35

118 54 109.3

4.2

7.8

149 8.0

1.5 130 27 77 3.4

0.06

0.09

0.02

45 7.50 22.13 3.44 17.58

47 7.34 20.80 3.27 17.34 5.42 1.64 6.04 1.14 7.49 1.60 4.71 0.74 4.53 0.67 3.88 0.87 0.46 0.17

150 50 165

5.50 1.73 6.48 1.20

7.78 1.68

4.94 0.78 4.79 0.71

4.16 0.48 0.64 0.17 0.689

49

4.8

361

222 96 97 94.5

81

1.2 128 31 103

3.8 0.19

6.9 0.05

126 29

16

12

19

3.34

3.16 9.74

4.54 13.32 2.06 10.84 3.56

10.41 1.66

9.26 3.15 1.15

3.76 0.73 4.56 0.96 2.98 0.45 2.76 0.41 2.03

0.19 0.16 0.08 0.401

1.57 8.71 2.95 1.06

3.50 0.65 4.29 0.90 2.74 0.42 2.58 0.39 1.91

1.02 0.18 0.07

1.25

4.25 0.78 5.21 1.12 3.25

0.50 3.10 0.45 2.58 0.56 0.39 0.23

62 73

102.5 23.3 235 43 148

27.9 0.67 99 21.40 43.15 5.27 22.48 5.42 1.76 6.00 1.03 6.46 1.38

4.05 0.62 3.68 0.55 3.80 2.02 1.67 0.53

121

Table 111.3 Major and trace element composition of DSDP Leg 29 basaltic basement cores. core

interval split analysis

Si02

Ti02 Al203 FeOt MnO MgO CaO

Na20 K20 P205 total

DSDP 282 DSDP 278 DSDP 279A DSDP 280A 278-35-02 278-35-03 280-23-02 282-20-02 282-20-01 279-13-02 107-113 80-85 64-67 112-117 48-54 106-110 glass w/r glass wir wir w/r xrl probe probe xii xr xr 50.18 48.65 50.43 50.95 48.68 49.89 1.48 1.50 1.42 0.90 1.01 0.91 16.27 17.83 15.44 18.19 15.66 19.04 8.50 8.83 9.18 8.23 7.24 9.49 0.14 0.16 0.14 0.10 0.15 0.13 5.86 10.47 8.13 8.73 7.96 8.36 11.75 7.65 11.95 11.50 13.39 13.29 3.11 3.51 2.31 2.46 3.01 3.13 0.07 0.09 0.19 0.10 0.18 0.22 0.13 0.12 0.10 0.08 0.16 0.06 99.35

100.87

100.91

98.59

40.9 252 430

33.3 197 317

42.0 316

36.8 213 293

140 100 69.8 1.8 125

115

69

76 55.1 3.1

59 72.9 3.9

125

160 23 80 12.5

98.78

DSDP 283 283-18-01 135-138 WI

xr 49.20 1.85 17.75 10.20 0.21 8.18 6.67

3.56 0.76 0.19

100.29

98.57

35.4 207 266 152

46.2 319

53

60 92.1 12.4 177 24 117 7.4 0.25 34 5.23 14.76 2.13 10.79 3.20

ICP-MS (ppm) Sc V

Cr Ni

Cu Zn Rb Sr Y

Zr Nb Cs

Ba La Ce

Pr Nd Sm Eu Gd

Tb Dy Ho

Er Tm Yb Lu Hf Ta Th U

Pb*

62 3.7 0.02

20 59 3.6 0.27

15

8

3.05 8.70 1.39 6.99 2.26 0.85 2.71 0.51 3.31

2.52 7.56 1.16 6.30 2.07 0.80 2.59 0.49 3.06 0.64 2.02 0.30

21

0.69 2.12 0.33 2.02 0.31 1.48 0.18 0.16 0.08

0.250

* isotope dilution analysis

1.81

0.28 1.35

0.24 0.18 0.07

178

0.06 65 7.17 17.54 2.37 11.75

159 122 65.8 1.2

90 18

45 1.5

101 3.1

0.22 20

0.24

1.23

3.13 10.88

4.19 0.69 4.22

3.22

1.58

1.20 3.74 0.67 4.18 0.88 2.64 0.41 2.43 0.38 2.23

0.69 2.17 0.43 2.98 0.65 2.03 0.33 2.16 0.33

0.80

0.16 0.15 0.04

0.71 0.23

70.0 4.2 162 30

1.25

12

1.83 10.00 3.24

261

59

1.21

1.21

3.82 0.70 4.66 0.98 2.90 0.45 2.73 0.40 2.29 0.22 0.15 0.05 0.423

3.62 0.67 4.23 0.88 2.63 0.40 2.58 0.38 2.87 0.46 0.41 0.18

122

Table P1.4 Major and trace element composition of dredge samples recovered near the Southeast Indian Ridge. AAD zone A MW2O-01 MW17-26 MW23-01 MW26-0l MW27-71 axis axis axis axis off-axis off-axis glass glass glass glass glass glass probe probe probe probe probe probe 51.34 51.1 52.39 50.14 50.87 50.62

MW8801

MWO7-01

dredge area split

analysis

Si02 Ti02

1.89 14.17 10.18 0.17

A1203

FeOt MnO MgO CaO

732

total

I

1.36 15.49 8.44 0.15

837

1.12 16.26 9.18 0.13 7.71 10.86 2.87 0.10 0.15

1.4 15.55

8.52 0.17 7.88 10.95 2.94 0.07 0.18

1.20 16.24 7.56 0.15 8.10 11.10 2.95 0.10 0.17

2.52 0.09 0.14

2.79 0.22

11.54 2.93 0.03 0.15

98.26

98.99

99.08 I

99.72

98.76

99.96

38.7 344 303 85 64 87.7 22.0 98

37.4 342 218

38.4 292 427

37.4 228 380

36.0 283 348

34.4 260

103 57

105

157

76 69.4 0.6

88

119 56

73.8 1.4

77.9 0.9

118 27

103 23

112 32

11.64

Na20 K20 P205

2.03 14.53 10.41 0.17 7.25 10.61 0.11

ICP-MS (ppm) Sc V

Cr Ni

Cu Zn Rb

Sr Y

Zr Nb Cs Ba La Ce

41 133

91.4 15.0 110 42 145

79

61

96

1.2 0.01

2.4 0.02

2.1

6

7

3.85 12.80

4.82

3 1.91

12.50

8.05

Pr Nd Sm

15.90 4.35

Eu Gd Tb

1.45

12.90 5.08 1.57

1.10

1.15

1.41 8.91

Dy

3.09 1.19 3.87 0.77 4.86

Ho

1.01

Er

3.06 0.49 2.88 0.43 2.22 0.08 0.07 0.02 0.226

Tm Yb Lu

Hf Ta Th

4.70 0.63 3.40 0.23 0.08

U

Pb*

* isotope dilution analysis

4.76 0.76 3.59 0.26 0.07

0.02

371 129 58

63.7 1.0 142 26 84

2.4 0.02

17

9

12

2.06 6.89 1.10

3.06 11.06

3.13 10.46 1.69 9.41 3.05 1.15 3.69 0.69 4.45 0.96 2.87 0.45 2.66 0.40 2.16 0.16 0.14 0.05 0.414

6.51

2.39 0.95 3.05 0.60 4.03 0.90 2.58 0.41 2.53 0.38 1.79 0.15

0.14 0.04 0.275

1.84 10.57 3.73 1.34 4.38 0.87

5.66 1.20

3.56 0.53 3.35 0.52 2.77 0.15 0.14 0.05

123

Table P1.5 Duplicate analyses of whole rock powder splits by XRF and ICP-MS. sample wir

ICP

V XRF

ICP

XRF

ICP

XRF

ICP

XRF

ICP

XRF

215

239

116

90

20

10

16

57

87

117

Sc

ICP XRF

Cr

Cu

Ni

Zn

264-15-cc

28

265-18-01

29

29

186

206

303

431

268

273

54

90

80

82

266-23-01

39

42

336

362

201

205

90

74

56

89

111

121

267-07-01

41

44

260

277

259

267

78

67

77

115

75

88

278-35-03

33

34

197

217

317

358

115

107

76

116

55

66

282-20-01

35

42

207

238

266

340

152

155

53

123

70

91

279-13-02

42

43

316

294

178

126

69

59

74

120

73

81

280-23-02

37

41

213

234

293

284

113

161

122

183

67

89

283-18-01

46

51

319

338

261

251

59

55

60

119

92

113

274-44-02

47

49

361

327

222

202

96

85

97

153

95

113

274-45-02

38

43

267

303

207

222

62

54

73

131

102

120

ICP

XRF

ICP

ICP

Y XRF

ICP

XRF

ICP

XRF

ICP

XRF

264-15-cc

20.3

19

205

210 24.6

30

111

148

7.2

8

311

292

265-18-01

7.2

8

257

283

25.3

27

112

121

13.2

13.4

97

65

266-23-01

4.4

5

129

140

49.2

50

149

154

8.0

8.9

46

17

267-07-01

4.7

4

126

132

28.7

29

81

80

3.8

3.1

11

0

278-35-03

3.0

3

125

131

20.1

21

59

62

3.6

4.3

6

0

282-20-01

4.1

3

162

176

29.9

32

101

102

3.1

3.3

11

0

279-13-02

3.9

6

166

164

20.2

24

80

87

12.1

12.1

56

280-23-02

1.1

2

90

101

14.4

19

45

54

1.5

2.4

20

6

283-18-01

12.5

14

177

191

23.7

26

117

117

7.4

9.2

33

27

sample w/r

30

Rb

Sr

XRF

Nb

Zr

Ba

274-44-02

1.1

0

128

134

31.4

36

103

104

6.9

8.8

18

0

274-45-02

23.6

25

235

274

42.7

44

148

149

27.9

26.6

98

75

124

Sample solutions for ICP-MS were prepared with -60-80 mg splits of glass and whole powders dissolved in tightly capped, 15 ml Savillex teflon beakers with -800 .t1 of a

HF:HNO3 acid mixture (1:3) heated at _800 C overnight. Upon dissolution, beakers were

uncapped and samples were dried on a hot plate to drive-off HF. Following drydown, the powders were then taken up once in 6N HC1 and twice in 4N HNO3, each dissolution step followed by an intervening drying step, to break down fluorosilicate precipitates. The final

dried powder was dissolved in 10 ml of 2N HNO3, from which a further 1:5 dilution was

prepared to run on the ICP-MS. A multiple, 20 ppb internal spike of Be, In, and Bi was

added to each sample to correct for machine drift. Unknown element abundances were determined using linear regression curves based on dissolved USGS rock standards (BIR1, BHVO-1, BCR-1, and W-1) processed along with the samples.

Sr, Nd, and Pb isotope analyses were performed on fresh glasses from DSDP sites

265, 266, 267, 278, and 282 and two off-axis dredge samples recovered directly east of the AAD (Table [V.6 and [V.7). Core samples without glass were subjected to a sequential

leaching procedure to remove alteration [Mahoney, 1987]. This leaching method differs from conventional warm acid techniques by being considerably more intense. A coarsely

powdered sample (200-800 mg) is leached in 4N to 6N ultra-pure HC1 and agitated ultrasonically for 20 minutes. The acid (clouded from the dissolution of alteration material)

is removed from the remaining solid and the procedure is repeated until the acid remains

clear during the leaching step. At this point, the final solid is removed, rinsed with ultra-

pure water and dried, leaving typically 10% to 40% of the original powered volume, depending on the extent of alteration. The leached powder is essentially composed of wellcrystallized plagioclase and clinopyroxene with very little, if any altered material [Mahoney,

1987]. Isotope dilution analysis of Nd, Sm, Sr, Rb, and Pb were determined on DSDP glass and leached whole rock powders, but not on dredge glasses recovered near the SEIR. Selected whole rock sample ages were determined by 40Ar-39Ar incremental heating

in order to clearly establish space-time relationships of the mantle sources in this region

125

Table IV.6 Sr, Nd, and Pb isotope ratios composition of DSDP Leg 28 cores. DSDP Leg 28 Cores DSDP site

L

264

265

266

267

274

274

07-01 78-82 glass 17.992 0.007 15.489 0.007 4.8 37.855 0.018

44-02 88-93 lchd w/r 18.800 0.005 15.499 0.005 -3.0 38.466 0.011

45-02

47.6 0.703039 0.000017

0.702696 0.000015

core interval split 2O6Pb/2O4Pb

15

17-01

23-1

cc lchd w/r 18.076

(±) 2O7Pb/2O4Pb

15.677

(±) 7/4 2O8Pb/2O4Pb

61-63 glass 18.080 0.007 15.497 0.006

22.7 39.175

pc. 5 glass 18.100 0.010 15.479 0.009 2.6 37.991 0.021 48.1 0.703083

0.713434 0.000016 0.511828 0.000010 -15.8

0.512893 0.000012 4.9

5.558

0.700

262.7

221.5

25.263

5.63 1

7.786 0.0012

11.562

(±)

L8/4 87Sr/86Sr (±)

LXSr 143Nd/144Nd (±) c Nd

Pb (ppm)

4.6 37.999 0.017 51.3 0.703490 0.000013

169.4

(±)

Sr (ppm)

0.689

0.401

110-1 14

Ichd w/r 19.484 0.003 15.590 0.003 -1.3 39.248

11.0

0.70289( 0.00001

0.250 0.0002

0.4

145.1

293

(±)

Rb (ppm) (±)

Nd (ppm) (±)

Sm (ppm) (±)

tNd, Sr and Pb isotopic values are not age corrected. ENd(T)=0 corresponds to i43Nd/i44Nd=0.5 12640. Isotopic fractionation corrections are 148NdO/l44NdO=0.242436 (l48Nd/l44Nd=0.24 1572), 86Sr/8Sr=0. 1194; Pb isotopes are corrected for fractionation using NBS 981 standard values of Todt et al. [1984]. Data reported relative to 143Nd/144Nd=0.51 1850 for La Jolla Nd, and 87Sr/8óSr=0.71024 for NBS 987 Sr standard. Within-run errors are less than or equal to the total ranges measured for La Jolla Nd (±0.000012; 0.2 c units), NBS 987 Sr (±0.000024), and NBS 981 Pb (206Pbf204Pb, ±0.010; 207Pbf204Pb, ±0.011; 208Pbf204Pb, ±0.038). Estimated uncertainties of isotope dilution analyses are